Cardiovascular disease remains the leading cause of death in the United States. Coronary magnetic resonance angiography (MRA) has the potential to become a non-invasive clinical alternative to diagnostic coronary catheterization, with an attendant reduction in both morbidity and cost. We propose that the combination of breath-hold 3D MRA, first-pass contrast enhancement, and partially parallel data acquisition, will provide unprecedented flexibility and robustness for coronary artery MRA, and help to establish it as a clinically useful test. The specific aims are: 1a) To implement breath-hold 3D MRA so that a volumetric study with a temporal window of 100 msec or less, and 0.5 mm or better in-plane spatial resolution, can be obtained. 1b) To implement the SMASH technique so as to improve temporal or spatial resolution in 3D MRA by a factor of 2-8. 2) To optimize the enhancement of the coronary arteries using contrast agents. To additionally test the feasibility of imaging the entire coronary tree in a single breath-hold using multiplanar, multiangle contrast-enhanced 3D MRA. 3) To evaluate the use of these techniques in a series of patients with documented coronary artery disease. The goal is to optimize the MRA method and to determine if the use of contrast agents provides any clinical benefit.